t. bone SC1100 aka Feilo Z3300 modding, anyone?

[micolas]

Might this be a tiny bit clearer?

http://recordinghacks.com/images/mic_extras/cad/gxl3000-schematic.png

Wow!
The scheme seems the same as Feilo,SE,T.bone,AdvancedAudio,etc.
I forgot CAD!
@Khron you have a fantastic memory!
(sorry, I'm on vacation, I can't think of schemes anymore)

 

[Khron]

The scheme seems the same as Feilo,SE,T.bone,AdvancedAudio,etc.
I forgot CAD!

And i also have a HiTec Audio "Fat Three", that's basically a rebadged GXL3000 (electronics AND body).

 

[micolas]

And i also have a HiTec Audio "Fat Three", that's basically a rebadged GXL3000 (electronics AND body).

20 years ago Gotham Audio from Canada marketed N3300 electronically
similar to SE Z3300 and T.Bone SC-1100.
Most likely they were all made by Feilo.

 

[Voyager10]

If you're putting a pad capacitor between the diaphragm and backplate (or diaphragm and ground, it acts the same, but will pop less when you switch it on), the value is: the attenuation factor you need, minus one, multiplied by the capsule capacitance.

A single sided LDC capsule is usually 60-70pF; if you wanted a 10dB pad that's a factor of 3, so the calculation gives you 120pF or 150pF. (If you're using both sides tied together, as in the GXL3000, the capacitance will be double).

For a pad capacitor between drain and gate you'll need to know what the drain-gate capacitance (Crss) is, which for the 2SK170 is quoted as 6pF but varies dramatically with drain-gate voltage. You'll probably want to try 10-22pF.

Make sure that the switch and all its wiring is on the 'ground' or 'drain' side of the cap, not the high-impedance side connected to the gate.

 

[Emmathom]

thanks @Voyager10 for the calculation method

so in summary, a 10-22pF between drain and gate (wire from switch to gate shorter as possible / capacitor on drain side of the switch) will do.
a x5,5 factor gives -15dB so [more or less] : 6pF*(5,5-1) = 27pF ? (got some polystyren 27pF)
I aim -15dB because the mic has a lot of output level...

well the right ratio is x0,18 to get -15dB so I did (1:0,18) = 5,5

 

[slipperyflax]

A couple of years back I did the mod suggested by Advanced Audio. It included capsule swap, transformer swap and one resistor change. Could go back and see exactly what it was if you are interested

 

[Voyager10]

thanks @Voyager10 for the calculation method

so in summary, a 10-22pF between drain and gate (wire from switch to gate shorter as possible / capacitor on drain side of the switch) will do.
a x5,5 factor gives -15dB so [more or less] : 6pF*(5,5-1) = 27pF ? (got some polystyren 27pF)
I aim -15dB because the mic has a lot of output level...

well the right ratio is x0,18 to get -15dB so I did (1:0,18) = 5,5

Yes, 27pF is about right, but be prepared to experiment.

The wiring should be

gate --- cap ---- switch --- drain

With the gate --- cap wiring kept short.

 

[Emmathom]

Yes, 27pF is about right, but be prepared to experiment.

The wiring should be

gate --- cap ---- switch --- drain

With the gate --- cap wiring kept short.

Thank you @Voyager10

And if we want to be a little bit more precise :

1) cap soldered on gate > short wire as possible > switch > short wire as possible > drain
or
2) gate > short wire as poss. > cap soldered on switch > short wire as possible > drain

physically the switch is like 3~4cm above the K170

 

[Voyager10]

(1), with the cap soldered onto the gate.

 

[abbey road d enfer]

The Lo-Cut switch in your case is necessary, because if you use it for kick-drum you must have a frequency response to the bass in the style of U47fet (which is notorious for kick-drum use)

I don't necessarily agree. In order to capture kick drum, you need first to address the level problem.
You have three solutions, for about 20dB attenuation:

• Brute force: put a 470-680pF capacitor in parallels with the FET input
• Chande to a charge amp: put a 68-100pF NFB capacitor between drain and gate of the FET
• Decrease bias voltage: choose a different tap on the voltage multiplier and/or use a voltage divider

I would opt for the latter.
Once the level issue is fixed, you may find that a HPF (bass-cut) is not mandatory, or can be applied later in the chain.
It may be counter intuitive, but there is not much proximity effect in a kick drum.
Proximity effect happens with point sources that generate spherical soundwaves.
A kick drum, because of the size and shape of its emitting source, generates almost flat soundwaves, which result in no proximity efffects.

 

[Emmathom]

I don't necessarily agree. In order to capture kick drum, you need first to address the level problem.
You have three solutions, for about 20dB attenuation:

• Brute force: put a 470-680pF capacitor in parallels with the FET input
• Chande to a charge amp: put a 68-100pF NFB capacitor between drain and gate of the FET
• Decrease bias voltage: choose a different tap on the voltage multiplier and/or use a voltage divider

I would opt for the latter.
Once the level issue is fixed, you may find that a HPF (bass-cut) is not mandatory, or can be applied later in the chain.
It may be counter intuitive, but there is not much proximity effect in a kick drum.
Proximity effect happens with point sources that generate spherical soundwaves.
A kick drum, because of the size and shape of its emitting source, generates almost flat soundwaves, which result in no proximity efffects.

in my case since the biasing is hot and therefore the output level is very high I would like a "pad" in order to capture strong sources rather close (such as for example a grand piano > mics at 20~40cm from strings)
1) Brute force: put a 470-680pF capacitor in parallels with the FET input : you mean in // with the 1nF already installed ?

2) ok for a cap. between drain & gate (what is a "NFB" capacitor ?) and according to @Voyager10 calculations it's more like 27pF~33pF (to experiment) to lower level by -15dB (which in terms of sound pressure only (not Fr) is equivalent to keeping away the mics by 2,5 times the distance - if I'm right)

3) Decrease bias voltage: choose a different tap on the voltage multiplier and/or use a voltage divider : don't we risk that this will bring noise (?)

4) what about adding a source R in // with the one installed ? the switch would work "reversed" since it's a ON/OFF
according to my schem. replace the 1,8KΩ by 3,9KΩ > switch ON = no pad (3,9KΩ // 3,9KΩ = 1,95KΩ = normal bias)
and switch OFF = pad (3,9KΩ = bias colder)

or a R serie (OFF = 1,8KΩ / ON = 1,8KΩ + ?KΩ)

 

[abbey road d enfer]

in my case since the biasing is hot and therefore the output level is very high I would like a "pad" in order to capture strong sources rather close (such as for example a grand piano > mics at 20~40cm from strings)

My Yamaha C7 is permanently equipped with 3 LDC's very close to the strings (about 10cm). I don't need a pad, they are set on 0dB attenuation.
Now if you want to use your mic in front of a Marshall stack, you need some attenuation.

1) Brute force: put a 470-680pF capacitor in parallels with the FET input : you mean in // with the 1nF already installed ?

No. One side to the capsule, the other to ground.

2) ok for a cap. between drain & gate (what is a "NFB" capacitor

Negative FeedBack. That's how charge amps work.

?) and according to @Voyager10 calculations it's more like 27pF~33pF (to experiment) to lower level by -15dB (which in terms of sound pressure only (not Fr) is equivalent to keeping away the mics by 2,5 times the distance - if I'm right)

The exact value depends on the actual capsule capacitance, the circuit's characteristics, and whatever amount of attenuation is deemed necessary. YMMV.
The attached picture seems to be for lesser attenuation and for a SDC.

3) Decrease bias voltage: choose a different tap on the voltage multiplier and/or use a voltage divider : don't we risk that this will bring noise (?)

Changing tap will not increase noise, on the contrary. Loading the DC/DC converter may create noise if the voltage divider draws too much current. I see that there are already a few 51Meg resistors that draw some current. If the voltage divider is properly implemented, that should not be an issue.

4) what about adding a source R in // with the one installed ? the switch would work "reversed" since it's a ON/OFF
according to my schem. replace the 1,8KΩ by 3,9KΩ > switch ON = no pad (3,9KΩ // 3,9KΩ = 1,95KΩ = normal bias)
and switch OFF = pad (3,9KΩ = bias colder)

This is playing with the op point.
When you "cold-bias", you lose headroom.
In order to get 15-20dB attenuation that way, you need to move the bias so cold you lose about the same amount of headroom, which leads you back to square one.

 

[Khron]

3) Decrease bias voltage: choose a different tap on the voltage multiplier and/or use a voltage divider : don't we risk that this will bring noise (?)

... But if the signal you're trying to capture is already too high in amplitude....... Is noise really an issue anymore?

 

[Emmathom]

My Yamaha C7 is permanently equipped with 3 LDC's very close to the strings (about 10cm). I don't need a pad, they are set on 0dB attenuation.

ok

No. One side to the capsule, the other to ground.

understood...

Negative FeedBack. That's how charge amps work.

yes but I didn't know the appellation "NFB"

The exact value depends on the actual capsule capacitance, the circuit's characteristics, and whatever amount of attenuation is deemed necessary. YMMV.
The attached picture seems to be for lesser attenuation and for a SDC.

which picture ? if it's mine it's a LDC t.bone SC1100
I understand this operation needs tests to find the right value from 27pF to maybe 220pF (?)

Changing tap will not increase noise, on the contrary. Loading the DC/DC converter may create noise if the voltage divider draws too much current. I see that there are already a few 51Meg resistors that draw some current. If the voltage divider is properly implemented, that should not be an issue.

This is playing with the op point.
When you "cold-bias", you lose headroom.
In order to get 15-20dB attenuation that way, you need to move the bias so cold you lose about the same amount of headroom, which leads you back to square one.

ok don't touch the R at the source
Thanks @abbey road d enfer for these explanations

 

[abbey road d enfer]

... But if the signal you're trying to capture is already too high in amplitude....... Is noise really an issue anymore?

What do you mean "too high"?
Te only thing that could be too high is if the diaphragm hits the backplate.

 

[Khron]

What do you mean "too high"?

"Too high" as in, clipping the electronics with the default (high) capsule bias voltage, nothing more - apologies for the lack of clarity (?)

 

[abbey road d enfer]

which picture ? if it's mine it's a LDC t.bone SC1100
I understand this operation needs tests to find the right value from 27pF to maybe 220pF (?)

I couldn'tmake it how much attenuation this is supposed to do...
Apparently it's not documented and I don't see a switch on the mic. Maybe internal...?

 

[abbey road d enfer]

"Too high" as in, clipping the electronics with the default (high) capsule bias voltage, nothing more - apologies for the lack of clarity (?)

The 3 solutions I proposed take care of signal at the source.

 

[Emmathom]

I couldn'tmake it how much attenuation this is supposed to do...
Apparently it's not documented and I don't see a switch on the mic. Maybe internal...?

The switch we intend to use is the low-cut one (and so forget about low-cut)
I'm refering to t.bone SC1100

 

[abbey road d enfer]

The switch we intend to use is the low-cut one (and so forget about low-cut)

OK. It wasn't clear.
IIUW the piece of schemo is a mod of an original, so the 27pF value is about random.